Karl W. Olson

On the steering of automated vehicles: Theory and experiment

Several facets of the automatic lateral control of individual ground vehicles are considered in detail. First, a path-dependent coordinate system for describing vehicle motion is defined, and the availability of motion quantities for control purposes is specified. Second, the lateral dynamics of a typical U.S. passenger sedan are empirically obtained and validated with data from full-scale studies. Third, various designs, in which different types of compensation are employed, are evaluated in terms of specified requirements and attractive candidates are specified. Finally, several controller designs were tested under full-scale conditions wherein a wire-reference configuration and a dual-mode test vehicle were employed. The latter was automatically steered on both straight and curving roads at speeds up to 35.8 m/s (80 mph). In one typical case, the maximum tracking error observed was 0.0635 m and occurred both when a sidewind was present and when the vehicle entered a curving section of roadway. Excellent lateral control-close tracking, good insensitivity to disturbance forces, and a comfortable ride-can be obtained using a relatively simple controller.

Use of Force and Attitude Sensors for Locomotion of a Legged Vehicle over Irregular Terrain

A number of legged vehicles are being developed for their mobility characteristics over irregular terrain. One such vehicle is The Ohio State University Hexapod vehicle (OSU Hexapod). Recently, the vehicle has been modified so that it can successfully walk over uneven terrain. Each of the feet has been equipped with two semiconductor strain gauges to measure lateral forces and a piezoelectric load cell to measure vertical forces. A vertical gyroscope and pendulums for orientation sensing have also been added. The control of locomotion over rough terrain can be decomposed into two complementary control processes, attitude control and active compliance. Attitude control is used to maintain the body tilt in a desired orientation, and active compliance is used to provide a suspension system by distributing force loading among the legs. It has been found that optimal force setpoints for the active-compliance algorithm can be calculated in closed form. This paper discusses the sensing hardware and the control system needed for legged locomotion over irregular terrain. Experimental results are provided for both statically determinate and indeterminate hexapod gaits.

A restructurable VLSI robotics vector processor architecture for real-time control

The authors propose a restructurable architecture based on a VLSI robotics vector processor (RVP) chip. It is specially tailored to exploit parallelism in the low-level matrix/vector operations characteristic of the kinematics and dynamics computations required for real-time control. The RVP is composed of three tightly synchronized 32-bit floating-point processors to provide adequate computational power. Besides adder and multiplier units in each processor, the RVP contains a triple register-file, dual shift network, and dual high-speed input/output (I/O) channels to satisfy the storage and data movement demands of the computations targeted. Efficiently synchronized multiple-RVP configurations, which may be viewed as variable very-long-instruction-word architectures, can be constructed and adapted to match the computational requirements of specific robotics computations. The use of the RVP is illustrated through a detailed example of the Jacobian computation, demonstrating good speedup over conventional microprocessors even with a single RVP. The RVP has been developed to be implementable on a single VLSI chip using 1.2- mu m CMOS technology, so that a single-board multiple-RVP system can be targeted for use on a mobile robot. >

Pipeline/Parallel algorithms for the jacobian and inverse dynamics computations

Algorithms have been developed for the Jacobian and Inverse Dynamics analyses in order to implement them on pipeline/parallel computing arrays. The results indicate that the sampling rate in either case may be significantly increased by adding processors to a pipelined array while, on the other hand, the compute time delay decreases very little. The results further show that a parallel structure is needed if the compute time is to be significantly reduced.

The electronic highway

Although high-speed, electrically powered surface and subsurface transportation will help to alleviate future traffic congestion in some localities, a majority of the public will probably continue to prefer the mobility, privacy, and freedom afforded by the automobile. Therefore, some form of automation for individual vehicles appears to be desirable-in fact, necessary- if complete highway chaos is to be avoided. The approach described in this article involves the concept of a dual-mode system, whereby the vehicle (which must be specially equipped) is manually controlled on nonautomated roads and automatically controlled on automated ones.

A VLSI robotics vector processor for real-time control

A VLSI robotics vector processor (RVP) for real-time control is described. Hardware parallelism and pipelining is used to exploit potential concurrency in the low-level matrix/vector operations characteristic of the kinematics and dynamics computations required for control. Three floating point processors (FPP), each with a adder, multiplier, and register file, all operating in parallel in an SIMD (single-instruction, multiple data stream) fashion are incorporated. Data exchange between the FPPs is facilitated by a dual shift-broadcast network. High-speed dual I/O channels are provided so that input/output bottlenecks are avoided. The RVP uses a RISC (reduced-instruction-set-computer) architecture with seven basic instructions. Composite vector operations such as matrix-vector multiply and vector cross product may be readily programmed using the basic instruction set to given considerable overlap. The RVP can be implemented on a single VLSI chip using 1.2- mu m CMOS.<<ETX>>

Design and implementation of a vision processing system for a walking machine

A vision processing system for a six-legged walking machine, the adaptive suspension vehicle, is presented. The vision-processing system consists of a laser range-finder, and vision computer, a terrain-elevation map, and a guidance computer. The range-finder measures the distances from itself to the objects in the scene. The specially designed vision computer processes the range data into a terrain-elevation form and stores the information with time data in a terrain-elevation map. With the real-time elevation information in the map, the guidance computer can select the best footholds for the walking machine in order to maneuver over rough terrain. >

Robotics vector processor architecture for real-time control

This paper proposes a restructurable architecture based on a VLSI Robotics Vector Processor (RVP) chip. It is specially tailored to exploit parallelism in the low-level matrix/vector operations characteristic of the kinematics and dynamics computations required for real-time control. The RVP is comprised of three tightly synchronized 32-bit floating-point processors to provide adequate computational power. Besides adder and multiplier units in each processor, the RVP contains a triple register-file, dual shift network and dual high-speed input/output channels to satisfy the storage and data movement demands of the computations targeted. Efficiently synchronized multiple-RVP configurations, that may be viewed as Variable-Very-Long-Instruction-Word (V2LIW) architectures, can be constructed and adapted to match the computational requirements of specific robotics computations. The use of the RVP is illustrated through a detailed example of the Jacobian computation, demonstrating good speedup over conventional microprocessors even with a single RVP. The RVP has been developed to be implementable on a single VLSI chip using a 1.2 μm CMOS technology, so that a single-board multiple-RVP system may be targeted for use on a mobile robot.

Skeletal motion processor for high speed robotics and graphical computations

A special-purpose Skeletal Motion Processor has been designed for real-time animation of human and animal figure motion. Several kinematic configurations for the lower torso of the human were investigated. The configuration which produced the best results consisted of two kinematically redundant manipulators, each having eight degrees-of-freedom, coupled by the sharing of two joints. The desired solution to the underspecified system of 12 equations and 14 unknowns was obtained using a pseudo inverse with an upper-level supervisory control function. As was expected, a real-time simulation of human skeletal motion was unobtainable with a general-purpose PDP 11/70 computer. With the Skeletal Motion Processor attached to a general-purpose computer, significant computation speed improvements are expected, thus enabling real-time simulation and control of human skeletal motion and robotic manipulator motion. The processor is capable of computing a Jacobian for a 7-link manipulator in 34 μs, and solving a system of six equations and six unknowns in 60 μs.

On Future Automated Ground Transport--Individual Vehicle Longitudinal Control

EXTRAIT La realisation d’un controle longitudinal sur et efficace est probablement le probleme technique le plus significatif en relation avec un vehicule automatique individuel, des systemes de transport tels que PRT, a double fonction, et avec la voierie publique automatique. Une structure de controle generale implequerait un controller central pour surveiller le fonctionnement du reseau, cela comprenant la coordination d’ordinateurs de secteur dont l’un superviserait et controlerait les vehicules fonctionnant dans le secteur qui lui est affecte. A ce niveau les trois facettes essentielles des operations sont: a) La determination de l’etat de chaque vehicule et quel rapport cet etat a avec l’etat desire; b) b) Les communications entre chaque vehicule controle et le controle de secteur; c) Le controle de chaque vehicule individuel. De recents progres concernant la conception, le developpement et l’experimentation de ces facettes seront decrits dans le contexte de la realisation d’un fonctionnement sur et efficace dans un court intervalle de temps (1-2 sec). Precisement, deux techniques ont ete mises au point en fournissant a un vehicule en marche des renseiqnements sur la position (ne permettant une marge d’erreur que de 20 cm ou moins), et sur la velocite (ou la marge d’erreur ne depasse pas 1.7%). D’autre part, un controller longitudinal de vehicule dont la marge d’erreur ne depasse pas 30 cm a l’arrivee de signaux de moindre et de grande intensite, reste insensible aux forces de perturbation et permet aussi aux passagers de jouir d’un trajet confortable, a ete concu, mis a l’epreuve et juge dans des conditions reelles de fonctionnaient.